Quad 303 Voltage Query

[fetteler]

Originally Posted by GlennJobson

OK, a couple of corrections from me.... Thanks for the help. I had misidentified a couple of transistors...

TR103 B 3.7 C 67.0 E 2.9

John BS - For instance, the collector of 101 must be the same potential as base of 102. Ditto, collector 103 = base 105 etc

That is the case, another correction

Collector 103 = Base 105 = 67.0V
Collector 101 = Base 102 = 1.4V

>The B-E voltage of Tr100 at 2.4V is also high...

TR100 B 4.5V C 1.9V E 5.1V

The diagram from Pete was super valuable.

That's better!!

Steve.

[PJL]

Originally Posted by Radio Wrangler

Damping factor is a terrible name. An output Z of zero or infinite Ohms can do no damping of speaker resonances whatsoever. Zero and infinity resistive impedances can dissipate no real power at all, therefore there can be no damping, no extraction of energy from resonances, no reduction of their Q.
David

The amplifier output stage absorbs the resonance energy. That said, the holy grail for ultra-low output impedance looks pretty stupid if you start looking at the cross-over design.

[Craig Sawyers]

OK - I understand most of the voltages and current around TR100, 101 and the base-emitter junction of TR102.

The it goes very screwy on the chain comprising R116, R117 and the Vbe multiplier.

Looking at the mulitplier, if the control is set midway, the voltage across it should be slightly more than 1.4V (it is 1.7V - so not far off). But TR102 collector (and hence base of TR104) should be close to 33.5V. It is actually 2V. Likewise the base of TR103 - likewise that should be close to 33.5V. It is actually 3.7V.

Thing is that the biassing of TR104 looks about right with 0.7V between base and emitter. The actual voltage is far wrong. But I'd say that TR102 and TR104 are both OK.

Things are very far wrong with the voltages around TR103, TR105 and or TR107 and or RV101. I'd concentrate on that bit of the circuit.

What is the voltage at the junction of R117, R117 and C106?

Good luck

Craig

[Radio Wrangler]

Originally Posted by PJL

The amplifier output stage absorbs the resonance energy. That said, the holy grail for ultra-low output impedance looks pretty stupid if you start looking at the cross-over design.

It's a loop circuit containing the amplifier output impedance, the cable impedance, the crossover circuitry and the driver voice coil impedances. The power dissipated in the series-connected items has to be proportional to their resistances.

If someone's got the amplifier output impedance lower than the overall speaker impedance, then most of the damping of resonance is within the speakers themselves. The lower the amplifier impedance, the less the amplifier's contribution to any damping, BUT this does better allow the other resistances to do the damping more effectively in what amounts to a series circuit.

This is a somewhat expanded version of what Craig said earlier.

David

[Craig Sawyers]

Originally Posted by PJL

The amplifier output stage absorbs the resonance energy. That said, the holy grail for ultra-low output impedance looks pretty stupid if you start looking at the cross-over design.

Suggest you read Richard Small's classic papers. All loudspeaker drivers are specified in terms of the Small/Thiele parameters (Thiele was Small's PhD supervisor).

But a snippet is that he analyzes a (low frequency) driver by converting it to an equivalent circuit (common practice).

The mechanical equivalent of the driver is driven by a Thevenin equivalent voltage source whose value is egBl/Sd x(Rg +RE). eg is the voltage from the amp, B is the magnet field strength, l is the length of wire in the gap, Sd is the radiating area, Rg is the output resistance of the amp and RE is the voice coil resistance.

Notice the Rg + RE on the bottom line? Provided that amplifier output resistance is significantly less than the voicecoil resistance, it can have no impact on the efficiency of the loudspeaker. 0.1 ohm is all that is needed

So what happens with the Q, which defines the mass spring resonance of the driver?

Well Small also, of course, analyzes that too, and comes to the conclusion that for any practical amplifier, where Rg is significantly less that RE there is no impact on overall system Q. That might be taken as less than 1% of the driver electrical resistance, or an output resistance for the amplifier of around 0.08 ohms. So again an output resistance from the amp of around 0.1 ohm is good enough.

Basically the voice coil resistance is by far the dominant factor in loudspeaker driver damping. The amplifier output resistance plays a tiny role in damping. The amplifier just does not absorb significant resonance energy by a country mile.

Craig

[PJL]

Agreed (with post #23), TR105 c & e readings are probably reversed.

TR103 B 3.7 C 67.0 E 2.9 looks suspect as it should be conducting hard but TR105 looks switched off. Try measuring the voltage on the junction of C106/R116/R117.

[Radio Wrangler]

The Quad 303 is a nice analogue, linear amplifier design. All transistors in quiescent condition should be biased into their active regions and carrying some current.

So a quick run around the Vbe voltages of them all ought to show them all in the 0.7v region.

Any showing noticably more points to a definitely faulty device.

Any showing noticably less means that either that device has gone short circuit, OR that whatever is feeding that base isn't doing so.

Once this is done looking at Vce voltages isn't quite so easy to interpret but can give more clues. Looking at voltages across resistors should reveal what currents are on the go, but this assumes the resistor is still OK.

It's a case of collecting clues and looking for trouble.

I find several prints of the schematic are helpful so I can write voltages across pairs of places and draw current arrows and values in places. If I find an inconsistancy and I've made an error or bad assumption, I'll need to start again on a fresh sheet. This way a picture starts to emerge of the extent of tha damage. Every part you can get any certainty over (Whether certainly good or certainly bad) is very valuable in reducing the areas of uncertainty, and it gives your mind a picture to work on. Lists of voltage readings don't let us use the pattern-recognition capabilities of our brains too easily. Things get clearer on pictures.

David

[PJL]

Here is one I did earlier...

[G4_Pete]

apologies mods for using yet more disc space on this system but this is now a clear scan of the previous images, I am not a computer whiz. A pdf gave better resolution that can be magnified without loss of detail.
Pete

[PJL]

Post #1 says TR1L base and collector are 67V but post #10 has TR1L B 3.0 C 67.0 E 2.8. Please check these again.

Also can you measure TR105 voltages but this time with the red probe connected to the 67V rail and not 0V.

Can you confirm all measurements were made without a speaker attached.

I still believe TR105 is not seeing any drive current into its base from TR103 indicating TR103 is faulty. The test above should confirm this.

Beware that the DC coupling may mean the failure was caused by a fault elsewhere!

A web search suggests replacements are BC546B or ZTX304, I would go for the ZTX which has a higher max. Vce (75V as opposed to 65V).

[Radio Wrangler]

Lets arbitrarily look at the top triplet first.

TR1L has low Vbe, it ought to be off. This fits with the output saturated near ground.

TR105 has low Vbe it ought to be off. This fits with TR1L etc.

TR103 has 0.8v Vbe, so ought to be on. This doesnt fit with TR105 onwards. TR103 is suspect open circuit. There is no current in R110. Double check measurements here. We may have our first duffer.

Now for the bottom triplet:

TR2L has 0.5v Vbe. Might be a little bit on depends on DVM accuracy. So the saturated output could be all due to the top triplet off. Still possible TR2L shorted but can't tell yet.

TR106 0.7v Vbe. should be plenty and should be turning TR2L on

TR104 -0.7v Vbe (PNP) should be plenty and turning TR106 on

So, the bottom triplet is pulling the amplifier output down. It couldbe OK and the issue is in a broken top triplet, probably TR103.

Now the VAS, voltage amplifier stage. TR102. No Vbe figure for this, but Vce is only 2v This part is highly stressed and second breakdown mode is a peril of VAS stages. Witout Vbe number not sure but TR102 is suspect ans worth testing/swapping.

It's time for some cheating!

Pull TR102 so you can test it without the distractions of the surrounding circuitry, but while it's out, fit a 22k resistot ro the board bridging the collector and emitter pads where TR102 has been removed.

Follow the path on the circuit down RV101, R116, R117, MR103, MR104 and the now added 22k. adding up the resistors and assuming RV101 is mid-way, then we have a voltage dropping chain of resistors and a couple of diodes from 67v su;ply to ground. This should bias the diodes somewhere near half of the 67v supply.

The triplets are essentially emitter followers. Lots of current gain but exactly unity voltage gain. so the amp output to the output blocker should be roughly 30v

We've spllt the amp in two, so we can test the output stages without the early stages, and without feedback. We can now debug and fix the output stages without distractions.

David

[GlennJobson]

OK, here we go....

Craig - C100 appears to test fine. I replaced C101 as I had another, though it tests OK.

Craig and PJL - R117 and C106 = 34.2V

G4 Pete - Thank you for the scan, very useful.

PJL - No speakers attached

PJL - Also can you measure TR105 voltages but this time with the red probe connected to the 67V rail and not 0V.

Tr105 = C 64.9 B 0.7 E 0.08

I have pulled Tr103, a ZTX304 and it appears to test as good.


Many thanks as always...

Glenn

[GlennJobson]

Also tested Tr102 out of circuit, ZTX304 tests as good.

TR2L is also testing as good.

[Craig Sawyers]

This is a thread worth reading https://www.diyaudio.com/community/t...torted.272430/

Also one channel distorted (left). Same or similar measured voltages as with yours. In his case it turned out to be R130 open circuit, and R104 had gone high - 83k when it originally was 22k. I think that R112 was also kaput.

Once he'd identified those and replaced two resistors, the left channel worked perfectly again.

Craig

[GlennJobson]

Originally Posted by Craig Sawyers

This is a thread worth reading https://www.diyaudio.com/community/t...torted.272430/

Also one channel distorted (left). Same or similar measured voltages as with yours. In his case it turned out to be R130 open circuit, and R104 had gone high - 83k when it originally was 22k. I think that R112 was also kaput.

Once he'd identified those and replaced two resistors, the left channel worked perfectly again.

Craig

Thanks Craig, I had found that post but of course I had not investigated enough, or had enough knowledge to know that that might be the issue I have here.

I will follow that up, hopefully tomorrow.

Glenn

[PJL]

TR105 Vbe appears to be forward biased and emitter current=8mA which is far too low. TR103 or possibly TR105 are now both suspects. TR103 base current and consequently TR105 base current is determined by R120 so it is worth checking it is 100 ohms and measure the voltage across R120 so we can work out the current.

[nutteronthebus]

There is a firm on ebay doing 303 driver boards for £34.95 may be a quick way to fix the problem https://www.ebay.co.uk/itm/134231976...kAAOSwCwhjaVMw

[GlennJobson]

Originally Posted by PJL

TR105 Vbe appears to be forward biased and emitter current=8mA which is far too low. TR103 or possibly TR105 are now both suspects. TR103 base current and consequently TR105 base current is determined by R120 so it is worth checking it is 100 ohms and measure the voltage across R120 so we can work out the current.

Not got to work on it today yet, I have though tried swapping the Tr105 from this driver with the one on the other board, same outcome. The L channel is distorted, the R channel is fine, so I think we can rule out the Tr105?

Glenn

[PJL]

Yes, TR103 remains the primary suspect but check R120 value and voltage first.

[PJL]

As the faulty channel is outputting something, this is not a 'hard' fault. Either biasing problem or transistor has low hfe and the focus is on the top triplet as that is not providing sufficient current.